Skip to main content
predict_game_teamwins() loads the saved model from models/nba_model.pkl and returns a win probability between 0 and 1 for a given set of rolling team statistics.
The model must be trained before calling this function. Run train_model_teamwins() first.

​
Function signature

predict_game_teamwins(
    points_diff,
    team_reb_roll,
    opponent_reb_roll,
    team_ast_roll,
    opponent_ast_roll,
    home
)
Returns: float — win probability between 0.0 and 1.0.

​
Parameters

​
points_diff
float
required
Rolling average points differential — the team’s average points per game minus the opponent’s, computed over the last several games.
​
team_reb_roll
float
required
The team’s rolling average rebounds per game.
​
opponent_reb_roll
float
required
The opponent’s rolling average rebounds per game.
​
team_ast_roll
float
required
The team’s rolling average assists per game.
​
opponent_ast_roll
float
required
The opponent’s rolling average assists per game.
​
home
int
required
1 if the team is playing at home, 0 if the team is playing away.

​
Usage examples

The examples below are drawn from main.py and test_betting.py.

​
Even match (from main.py)

A team with no points advantage, slightly weaker rebounding and assists, playing at home: 
from prediction_ai import predict_game_teamwins, evaluate_bet_teamwins

prob = predict_game_teamwins(
    points_diff=0,
    team_reb_roll=4,
    opponent_reb_roll=5,
    team_ast_roll=4,
    opponent_ast_roll=5,
    home=1
)
print("Win Probability:", round(prob, 3))
print("Evaluation:", evaluate_bet_teamwins(prob))

​
Strong home favorite

A team with a clear statistical edge and home court advantage: 
prob = predict_game_teamwins(
    points_diff=10,
    team_reb_roll=5,
    opponent_reb_roll=3,
    team_ast_roll=4,
    opponent_ast_roll=2,
    home=1
)

​
Slight underdog

A team trailing in most metrics and playing away: 
prob = predict_game_teamwins(
    points_diff=-2,
    team_reb_roll=4,
    opponent_reb_roll=5,
    team_ast_roll=3,
    opponent_ast_roll=4,
    home=0
)

​
How it works

Internally, the function constructs a single-row DataFrame from the six inputs, then calls predict_proba on the loaded model: 
def predict_game_teamwins(points_diff, team_reb_roll, opponent_reb_roll,
                          team_ast_roll, opponent_ast_roll, home):
    model = pickle.load(open(MODELPATH, "rb"))
    features = pd.DataFrame([{
        "points_diff": points_diff,
        "team_reb_roll": team_reb_roll,
        "opponent_reb_roll": opponent_reb_roll,
        "team_ast_roll": team_ast_roll,
        "opponent_ast_roll": opponent_ast_roll,
        "home": home
    }])

    prob = model.predict_proba(features)[0][1]
    return prob
predict_proba(features)[0][1] returns the probability assigned to class 1 (win).

Build docs developers (and LLMs) love

Get started for freeTalk to us